Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: My reason for wanting this bug fixed.

Author: Vincent Diepeveen

Date: 04:02:08 04/04/01

Go up one level in this thread


On April 02, 2001 at 19:58:05, Uri Blass wrote:

>On April 02, 2001 at 17:57:53, Laurence Chen wrote:
>
>>On April 02, 2001 at 12:17:44, James T. Walker wrote:
>>
>>>On April 02, 2001 at 09:32:59, Laurence Chen wrote:
>>>
>>>>On April 01, 2001 at 23:12:07, James T. Walker wrote:
>>>>
>>>>>On April 01, 2001 at 22:13:01, Laurence Chen wrote:
>>>>>
>>>>>>Is it me or other users are also experiencing the same "bug"?  I've just updated
>>>>>>Gambit Tiger to the new version, and the first thing I've tested was the
>>>>>>analysis mode, and to my disappointment it fails to analyze if the position is
>>>>>>an endgame if there's 5 pieces or less.  I have all the 5 men tablebases, and
>>>>>>the analysis mode fails to suggest any move at all.  I think the Beta testers
>>>>>>failed to check for bugs and they were supposed to check for all bugs, no for
>>>>>>chess engine strenght alone, this annoys me.  Well I hope that Christopher can
>>>>>>fix this minor bug.
>>>>>
>>>>>Hello Laurence,
>>>>>Why do you want to "analyze" a 5 man endgame when the tablebases will give you
>>>>>the best move?  I guess you are right, it's a bug but I don't see it as major
>>>>>since just telling it to move will give the "best move".  Also if you turn off
>>>>>the "Use endgame databases" option under the "Extra/options" menus it will
>>>>>analyze the endgame for you.
>>>>>Jim
>>>>Hi Jim,
>>>>
>>>>It's not that I want to analyze a 5 men endgame, it surprised me that the
>>>>analysis mode failed when I loaded any 3, 4, or 5 men endgame.  Of course, I
>>>>could use any of my Chessbase engines in order to analyze 5 men endgame.  Put
>>>>this way, what if I set up a position with 7 men endgame, and it gets to a point
>>>>in which a trade of pieces is possible to reduce to a 5 men endgame, because
>>>>Chess Patner only displays the principal PV, unlike Chessbase GUI which gives
>>>>all other alternatives, I like to try different moves in the analyzis mode and
>>>>see what PV score GT or CT would give.   And if the endgame gets reduced to a 5
>>>>men endgame, I won't be able to see the evaluation because of this bug.  That
>>>>means that I would have to use a Chessbase engine to analyze the position, which
>>>>means closing CT or GT and open another program.  Then what would be point of
>>>>using GT or CT to analyze endgame positions if I need to open another chess
>>>>program to do a follow up. Grrr....
>>>>Regards,
>>>>Laurence
>>>
>>>Hello Laurence,
>>>I still do not understand your problem.  As I said above if you want to analyze
>>>the position you can turn off the Tablebases then Tiger will analyze per normal.
>>> If you are in the 5 man tablebases then what analysis is needed?  The tablebase
>>>gives you the answer to any move you input.  So you can analyze with tablebases
>>>on untill you get to 5 men.  At that point you can get the "definitive" answer
>>>to any move from the tablebases or you can turn off the tablebases and let Tiger
>>>analyze without tablebase info.  What could you miss with this procedure?  With
>>>tablebases on you get the "perfect" solution and with them off you get Tiger's
>>>opinion (not necessarily perfect).
>>>Jim
>>Hi Jim,
>>You and I are blessed to the fact that we can afford to buy new chess engines
>>updates when they come out.  What if you did not have any chess engine(s) which
>>use(s) the tablebases?  Because you have other chess engines which can access
>>the endgame tablebases, you feel that this is not necessary.  I wonder if you
>>would feel differently if this was the only chess engine which you could have. I
>>bet you would want to see the truth in 5 men endgames.
>>Let's say that this was your only chess engine you could buy, and you wanted to
>>study Rooks Endings (Rook+Pawn+King vs. Rook+King), and you wanted to know the
>>proper technique of playing these type of positions, and if the chess engine
>>cannot display the moves how would you know what the proper technique.
>
>It is more convenient to look at the analysis mode but you can still learn
>without other programs from asking tiger to play.
>
>  Not
>>everyone can afford to buy many chess engines, and some of us like to use chess
>>engines to learn technique in chess.
>
>A small minority use engines to learn playing tablebases positions.
>Even if you use engines as a training tool there are more important things than
>tablebase.
>
>I think that tablebases are not relevant in 99% of the human-human games.

If you don't mind i raise that to 99.9%
and that includes using them in the search before you get to EGTBs.

I had one KQP KQ back in 1995 or whatever with me on the Q side,
opponent had a-pawn.

It appears i drew it very efficiently, but that at the moment that the
endgame started the EGTB says it was a win for my opponent.

However i see EGTB relevance for blitz very clearly, whereas for computer
chess the relevance of egtbs is far more important as in 1% of the games.

I see so many endgames where it helps at *any* level!

>Uri



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.